Materialization
SDF supports materializing tables in Snowflake and Redshift directly from the command line.
Overview
Runtime materialization helps to bring your SDF model from a development environment to live tables in production. It works against remote warehouses and the local SDF database alike:
- Materialize models as views, tables, incremental tables, transient tables, temporary views, recursive tables, and more.
- Debug your models faster and at no cost, since static analysis checks errors before running them.
- Run tables with Jinja, SQL Variables, or other relevant metadata without manually converting them to plain SQL.
Materialization is executed as part of sdf run.
Materialization Prerequisites
Prerequisites
In order to materialize tables in your data warehouse, you will need to have the following:
- A Snowflake or Redshift warehouse to materialize tables remotely.
- Valid credentials with write access to materialize tables in your warehouse.
- A configured integration in your workspace.sdf.yml file.
As such, we recommend starting with our integrations documentation before proceeding with materialization.
Materialization Guides
Below are a list of follow-along guides to help you get started with materialization in SDF. Since materialization differs between warehouses, each guide walks you through a specific materialization type in a specific warehouse.
Materialization Type | Snowflake | Redshift | BigQuery |
---|---|---|---|
View | Link | Coming soon… | Coming soon… |
Table | Link | Coming soon… | Coming soon… |
Incremental | Link | Coming soon… | Coming soon… |
SDF supports a wide variety of materialization options. See the materialization config to view them all.